Geographically Diverse Storage Data Convolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage techniques for geographically diverse storage systems face inefficiencies due to high disk read/write events, network traffic, and resource consumption from redundant data replication, which can lead to increased storage space usage and energy costs.
Innovation Solution
Implementing data convolution using XOR operations or other logical/mathematical operations to combine data chunks from different zones into a single, compressed chunk, reducing the need for redundant copies and minimizing data transfers, while allowing for efficient recovery of original data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant data replication is used for geographically diverse storage, then data reliability is improved, but storage space usage and resource consumption increase
Solution Approach 1:
The patent combines multiple data chunks from different geographic zones into a single convolved chunk using XOR operations. Instead of storing separate redundant copies of data from zone 1, zone 2, and zone 3, the system merges them into one convolved chunk that contains encoded information from all zones, reducing total storage space while maintaining data reliability through the convolution relationship
Solution Approach 2:
The patent transforms data from its original redundant replication form into a convolved form by changing the storage parameter from separate identical copies to a mathematically transformed single copy. The XOR operation changes the data representation while preserving the ability to recover original data, thus reducing storage quantity while maintaining reliability
2Reliability
If redundant data replication is performed across geographically diverse zones, then data redundancy is improved, but network traffic and resource consumption increase
Solution Approach 1:
The patent merges the function of creating multiple redundant copies with a single convolution operation. Instead of separately replicating data to multiple zones and transmitting it across the network, the system performs a single convolution operation that simultaneously creates the redundant relationship across zones, reducing network traffic and computational resource consumption
Solution Approach 2:
The patent extracts the redundancy creation process from traditional replication and replaces it with a convolution operation. By taking out the separate replication steps and replacing them with a single XOR-based convolution, the system eliminates unnecessary network traffic and resource consumption while maintaining the essential redundancy function
3Reliability
If data chunks are replicated to geographically diverse locations, then disaster recovery capability is improved, but disk read/write events increase
Solution Approach 1:
The patent merges multiple data writing operations into a single disk write operation. Instead of writing separate redundant copies to different disks in different zones, the system convolves the data chunks and writes a single convolved chunk, dramatically reducing the number of disk write events while preserving disaster recovery capability through the convolution relationship
Data Source
AI summary
Data convolution for geographically diverse storage is disclosed. Data chunks stored in storage devices of different zones of a zone storage system can be convolved to conserve memory. The zone storage system can be a geographically diverse storage system. A convolved chunk can be stored at a zone that does not contribute to the data represented in the convolved chunk. In an aspect, a copy of a first chunk from a first zone can be created at a third zone, which can facilitate convolving the data of the copy with data from a second zone as the data from the second chunk of the second zone is received at the third zone, e.g., without having to create a local copy of the second chunk, which can be termed ‘on-arrival convolving’ or other similar terms. The copy of the first chunk can then be deleted. Copies of the convolved chunk can be created.